The Evidence for Demons

This post isn't actually arguing for the existence of demons; it is actually about the misuse of neuroscience. I was reading Oliver Sachs new book, and he said something (the book is not with me now) along the lines of, "Some people took these feelings of presence as evidence of demons or ghosts, but now we know they are caused by such-and-such neural phenomena."

How in the world does that dispose of the possibility of demons or ghosts existing? (Again, please, this is not an argument for their existence: I am addressing neuroscientists trying to do philosophy, not the spirit world!) Don't, say, trees create particular neural patterns when we look at them? Does that prove that trees aren't real as well? Why can't someone who believes in ghosts say, "See, now we have physical evidence for ghosts: they are able to create those neural patterns you see!"

Look, when it comes to brain functioning, neuro-scientists are beast. (Yeah, I meant to write "beast": it's the way my 12-year-old describes someone who "rules.") When they start doing metaphysics, they are in no better position than plumbers or bridge engineers: a neuro-scientist might coincidentally be a good metaphysician, but if so, it is not because he knows a lot about the brain!


  1. Replies
    1. Why would you think a post making a simple philosophical point is a "swipe" at anyone?

  2. Maybe I'm reading too much DeLong, but sometimes it seems you and Feser are the only ones in the world pointing out these obvious problems.

  3. Because no-one can find any evidence of ghosts EXCEPT these "neural patterns." This is just another example of the problem of induction, and the answer is the same: judge by Occam's Razor and predictive success. Is that a complete answer to the problem of induction? Of course not. It's the best there is, that's all.

    1. Ken, Ken, Ken: I very specifically said I don't want to argue the existence of ghosts here!

      The question I do want to argue is this: You have some probability you give to the proposition "Ghosts exist": call that probability x. Now Sachs comes along and says, "When people say they are seeing ghosts, we find this specific neural pattern occurring."

      Should this fact lower x for you if you are a good Bayesian updater? Sachs seems to say "yes," but I say that is nonsense: for any value of x, I would think the person believing that value would *already* think "ghosts" create some neural pattern. So, for instance, if Joe "Ghost Hunter" Smith has an x of 99% before he reads Sachs because he strongly believes ghosts are real, is there any reason he should change that value? Why wouldn't he think, "Well, yes, just like my wife coming in the room creates a neural pattern, so does a ghost coming in the room!"

      Please, please don't start arguing with me that there is no reason for Joe to assign a 99% probability to ghosts existing: I am asking should be the impact of Sachs's "new evidence"?

    2. But that's not Sachs's argument. He's not saying we observe such a pattern whenever we have a ghost citing. We don't trail Bill Murray when he's on the job ghost-busting. Sachs is citing this as part of a wider phenomenon. Take phantom limbs. People feel pain in legs they no longer have; the best explanation is neural activity in the brain. Same he says with ghosts. It's not proof ghosts don't exist, just like and understanding of thunder is no proof Thor doesn't exist. It's an argument that we needn't posit ghosts, or Thor. "Sire I had no need of that hypothesis" is a dismissal not a disproof.

    3. Ah! And so I will henceforth dismiss trees in just the same way: after all, all of the evidence I have for them is just "neural activity in my brain"!

    4. And by the way, Ken, have you actually read Sach's latest book, or are you just making up a version of what his argument is?

    5. I have not read a Sachs book in some years Gene. I read a bunch in the 90s. I am trusting you to have quoted fairly and accurately.

    6. Ken, you are both trusting my version of Sachs AND saying I have gotten his argument wrong?!


Post a Comment

Popular posts from this blog

Central Planning Works!

Availability bias